jax-ml/jax · error · TypeError

iteration over a 0-d array

Error message

iteration over a 0-d array

What it means

EArray (an etuple-backed lazy array used internally by JAX's symbolic/OOO machinery) mimics ndarray semantics; NumPy disallows iteration over 0-d arrays, so EArray.__iter__ raises TypeError for ndim==0 and NotImplementedError otherwise.

Source

Thrown at jax/_src/earray.py:58

  @property
  def aval(self):
    return self._aval

  def block_until_ready(self):
    _ = self._data.block_until_ready()
    return self

  def copy_to_host_async(self):
    self._data.copy_to_host_async()

  def copy(self):
    return EArray(self.aval, self._data.copy())

  def __repr__(self):
    return 'E' + repr(self._data)

  def __iter__(self):
    if self.ndim == 0: raise TypeError('iteration over a 0-d array')
    raise NotImplementedError

  # forward to aval
  shape = property(lambda self: self.aval.shape)
  dtype = property(lambda self: self.aval.dtype)

  # computed from shape and dtype
  ndim = property(lambda self: len(self.aval.shape))
  size = property(lambda self: math.prod(self.aval.shape))
  itemsize = property(lambda self: self.aval.dtype.itemsize)
  def __len__(self):
    if self.ndim == 0: raise TypeError('len() of unsized object')
    return self.shape[0]

  # forward to self._data
  devices = property(lambda self: self._data.devices)  # pyrefly: ignore[bad-override]
  _committed = property(lambda self: self._data._committed)
  is_fully_addressable = property(lambda self: self._data.is_fully_addressable)

View on GitHub (pinned to 1e1c6a8fc0)

Solutions

1. Check ndim/shape before iterating: if earr.ndim > 0: ...
2. Index the scalar directly: earr[()] to extract the 0-d value
3. Avoid generic iteration over unknown JAX container types; convert to a concrete array first

Example fix

# before
vals = list(earr)  # TypeError when 0-d

# after
vals = [earr[()]] if earr.ndim == 0 else list(earr)

if earr.ndim == 0:
    value = earr[()]  # extract scalar instead of iterating

def is_iterable_earray(e) -> bool:
    return getattr(e, 'ndim', 0) > 0

When it happens

Trigger: Iterating (for x in earr, list(earr), *earr) over an EArray whose aval shape is (). This class is internal to jax._src.earray and mostly encountered when embedding etuples in control-flow/metaprogramming code.

Common situations: Internal JAX development or libraries built on jax.experimental serialization / etuples hitting an EArray where a scalar was expected; debugging repr-based code that iterates values.

Related errors

• len() of unsized object
• iteration over a 0-d array
• len() of unsized object
• iteration over a 0-d key array
• numpy masked arrays are not supported as direct inputs to JA

AI-assisted analysis of jax-ml/jax@1e1c6a8fc0 (2026-08-27). Data as JSON: /api/errors/93b6c7a4ea933b64. Report an issue: GitHub.